HP RF Design and Measurement Seminar

1. HP RF Design and Measurement Seminar Creators: David Ballo Andy Potter Boyd Shaw My Le Truong Joe Civello Ed Henicle Sara Meszaros

2. Goals of the Seminar • Introduce new engineers to the modern RF design process • Document predictive RF design process • Focus on design methodology, not button pushing • Provide practical design tips based on our case study

3. Agenda • Introduction • Overview of RF Design Process • Case Study: RF Front-End • Low-Noise Amplifier (LNA) • Duplexer • Power Amplifier • Measurement for Design • Passive Device Characterization • Active Device Modeling • Summary

4. Traditional RF Design Process • Paper, pencil, calculator • Many board turns (cut and try) • Long design cycles

5. Modern, Predictive RF Design Process • Combination of: • EDA (electronic design automation) software • Measurement equipment (e.g., network & signal analyzers) • Design iterations now performed via software • Fewer board turns (faster time to market) • Accurate circuit performance - minimize over engineering • Improve manufacturability with yield analysis and optimization

6. Agenda • Introduction • Overview of RF Design Process • Case Study: RF Front-End • Low-Noise Amplifier (LNA) • Duplexer • Power Amplifier • Measurement for Design • Passive Device Characterization • Active Device Modeling • Summary

7. Three Critical Design Considerations Time to Market Performance Cost

8. General RF Design Process Concept Design Integration

9. Concept:System Design\Analysis\Partition • Assess goals • Set priorities • Explore possible system configurations • Design partitioning • Allocate circuit specifications

10. Id h Vds w Circuit Level Design • Explore possible circuit topologies • Investigate and select components • Make build vs. buy decisions • Determine whether desired circuit Specifications are realizable • Verify system performance

11. Integration • Combine individual circuits • Troubleshoot • System redesign, as needed • Circuit redesign • Circuit reallocation • System reconfiguration • Modify system specification, as needed • Re-define project definition, as needed

12. The Traditional Design Process

13. Concept:System Design/Analysis/Partition • Understand needs and goals • Paper study • Prone to errors • Incomplete system analysis • Difficult to analyze circuit interactions

14. Circuit Level Design • Reliance on physical breadboards (prototypes) • Trial-and-error circuit design (cut & try) • Difficult to verify circuit's performance on system level • Difficult to predict interactions & manufacturing yield • Expensive & time consuming!!!

15. circuits system Once integrated, system = circuit + circuit interactions interactions Integration & Test • Trouble shooting • System level • Circuit level • Little flexibility • Expensive changes • Inefficient process spurs! oscillation! I didn't see any of this coming! When do I go home?

16. Traditional Design Process Design Concept Integrate Redesign Design Build Integrate Test Does it work? Production NO Production

17. Modern, Predictive RF Design Process

18. Concept Design Integration Modern, Predictive RF Design Process • Combine test equipment and EDA software for fast, efficient design! Concept Design Integration

19. Investment in Models • Dedicate modeling team • Use available parts libraries • Measure individual parts • Different operating conditions • Improve individual parts models with time Ibe=(IBbif(exp(Vbe/NbfVT)-1.0))+Ise(exp(Vbe/(NexVt))-1.0) !Freq.[Hz] MagS11[dB] PhaseS11[DEG] MagS21[dB] PhaseS21[DEG] MagS12[dB] PhaseS12[DEG] 300000 -5.986E-07 -1.151E-02 -7.394E+01 8.997E+01 -7.394E+01 8.997E+01 -5.986E-07 -1.151E-02 315229 -6.384E-07 -1.210E-02 -7.351E+01 8.997E+01 -7.351E+01 8.997E+01 -6.384E-07 -1.210E-02 331231 -6.812E-07 -1.271E-02 -7.308E+01 8.997E+01 -7.308E+01 8.997E+01 -6.812E-07 -1.271E-02 348046 -7.273E-07 -1.336E-02 -7.265E+01 8.997E+01 -7.265E+01 8.997E+01 -7.273E-07 -1.336E-02 365714 -7.769E-07 -1.403E-02 -7.222E+01 8.997E+01 -7.222E+01 8.997E+01 -7.769E-07 -1.403E-02 384279 -8.303E-07 -1.475E-02 -7.179E+01 8.997E+01 -7.179E+01 8.997E+01 -8.303E-07 -1.475E-02 403787 -8.879E-07 -1.550E-02 -7.136E+01 8.997E+01 -7.136E+01 8.997E+01 -8.879E-07 -1.550E-02 424285 -9.501E-07 -1.628E-02 -7.093E+01 8.997E+01 -7.093E+01 8.997E+01 -9.501E-07 -1.628E-02 445823 -1.017E-06 -1.711E-02 -7.050E+01 8.997E+01 -7.050E+01 8.997E+01 -1.017E-06 -1.711E-02 468455 -1.090E-06 -1.798E-02 -7.007E+01 8.997E+01 -7.007E+01 8.997E+01 -1.090E-06 -1.798E-02 492235 -1.168E-06 -1.889E-02 -6.964E+01 8.997E+01 -6.964E+01 8.997E+01 -1.168E-06 -1.889E-02 517223 -1.252E-06 -1.985E-02 -6.921E+01 8.997E+01 -6.921E+01 8.997E+01 -1.252E-06 -1.985E-02 543479 -1.344E-06 -2.086E-02 -6.878E+01 8.997E+01 -6.878E+01 8.997E+01 -1.344E-06 -2.086E-02

20. Quickly and accurately analyze system performance ! Concept: System Design/Analysis/Partition circuits system • Top-level system brainstorming • Quick analysis ofcircuit interactions • Budget analysis to allocate circuit specifications • Design partitioning system-level interactions

21. Block & Level System Budget Analysis

22. Classic Example: Spur Analysis • Select frequency plan based on predicted performance

23. Receiver Block D/A I AGC A/D Baseband DSP Frequency Synthesizer 90 A/D Q From Synthesizer Circuit Envelope Simulation Transient Simulation HP Ptolemy Simulation Integrated Design Environment To D/A

24. Design Modify design to match proto Simulate (Circuit & System) Good Build physical proto & test Good Integrate Circuit Design Cycle

25. Circuit DesignBrainstorm • Experiment with possible biasing schemes • Explore different circuit configurations • Make many different analyses quickly

26. Continuing to optimize with- out a physical Careful! Proceed with caution. prototype Why Modify Design to Match Prototype? First, verify that there are no errors in the fabrication or measurement of the prototype. Why modify the design to match the prototype? Provides starting point to refine design Gives good assurance that design changes will indeed improve design direction! The closer prototype matches the model, the greater the probability for success : right c Ps =(k) x 100 where Ps= success probability c = # of changes k = guru factor (<1.0)

27. s21 log mag 10dB/div 0dB ref simulated measured 2 Integration: Layout & Prototype • Interconnect designs and measure performance • Modify design to reflect measured results (if necessary)

28. Measured System Modeled System Production Success! Measured = Modeled & Exceeds Spec

29. Design Integrate Software Simulate Redesign Does it work? NO Manufacturable? NO Build Test NO Does it work? Physical Production The Predictive RF Design Process Concept • Tweak the design on the work station, not on the bench! Design Integrate Production

30. Solutions from Hewlett-Packard

31. HP Advanced Design System Total integration of schematic, simulation, and layout

32. The RF Design Process Concept Design Production Integration & Test • System Analysis • Design Partitioning • RF • Analog • DSP • Integrate Blocks • System Measurements • Re-Layout • Final Artwork • Bill of Materials • Documentation • Integrated Simulators • Faster Simulators • Optimizers • Instrument I/O • Parts Libraries • Co-Simulation • System Simulation • Layout • EM Simulation • Parts Libraries • Third Party Links • Artwork Generation

33. HP’s Measurement Instrumentation • HP Advanced Design System • Network and Signal Analyzers • Protocol Signal Sources (CDMA, GSM, etc.)